Namespace:

Canonical resource grammar: me://namespace[selector]/path Implementation-specific daemon command surface: me://namespace:operation/path 1 key = 1 subject = N monads.ai namespace = canonical resource namespace selector = canonical NRP selector when using the resource grammar operation = daemon/kernel command intent when using the command surface path = semantic route or command target

El claim ahora está anclado al identity hash del kernel en lugar de derivarlo de namespace + secret.

The recommended canonical resource form is the one specified by the Namespace Resolution Protocol. Command-style me://namespace:operation/path targets remain part of the current daemon implementation surface, but they do not replace the canonical NRP grammar.

me://self:read/profile
me://self:write/profile.name
me://self:inspect/profile
me://self:explain/profile.netWorth
me://kernel:read/memory
me://kernel:export/snapshot
me://kernel:import/snapshot
me://kernel:replay/memory
me://kernel:rehydrate/snapshot
me://kernel:get/recompute.mode
me://kernel:set/recompute.mode
me://kernel[device:localhost|protocol:http|port:8161]:export/snapshot
me://ana[host:foo|protocol:https]:read/profile
me://wikileaks[host:wikileaks.org|protocol:https]:read/page

1. Explainability: me://self:explain/profile.netWorth is readable by humans and machines. It’s a request for a logic trace on a specific data point.
2. Remote Kernel Control: me://kernel[host:office]:rehydrate/snapshot allows you to manage a remote daemon with the exact same syntax as a local one.
3. Clean Ledger: Every entry in the monad.ai ledger will now follow a strict Subject -> Action -> Object structure, making audits and replays trivial.

So now the current daemon command surface reads cleanly: me://self:read/profile me://self:write/profile.name me://kernel:export/snapshot me://kernel:replay/memory

Algebra of Encrypted Audiences

In this algebra, access and visibility are determined by the interplay of three sets:

• Context defines where and on which nodes data resides.
• Capability defines which actions are permitted on the data.
• Encryption defines who can decrypt and read the content.

Mathematically, we express these properties as three sets:

• T = Topology set
  • On which nodes does the ciphertext exist or replicate?
• A = Audience set
  • Which identities are authorized to decrypt?
• P = Procedure/Capability set
  • What actions can be performed on this content?

Thus, an “encrypted island” can be formally represented as:

Island I = (path, ciphertext, T, A, P)

Example:
Suppose an image is saved and replicated across:

• office-node
• iphone
• backup-daemon

Topology set:

T = {office, iphone, backup}

But only you and your spouse can decrypt it:

A = {me, wife}

In other words:

• Multiple nodes may store the island, but only a specific subset of users can read it.

A clean way to conceptualize this is:

• “Space” is the primary entity: all other properties are predicates or projections over that space.
• Everything else—membership, replication, visibility, or capabilities—is expressed as refinements, unions, intersections, or other set operations.

Common cases, then, are simple statements about the sets:

You can formalize any “space” as:

• A: The audience set (who can read)
• T: The topology set (where it lives)
• C: The capability/action set (what can be done)
• P: The path/subspace (context or semantic location)

Adjectives like “private”, “shared”, “replicated”, or “encrypted” are then just properties over those sets. For example:

• private: Only the owner can read (A = {self})
• shared: More than one party can read (|A| > 1)
• public: Anyone can read (A is open)
• replicated: The content exists on multiple nodes (|T| > 1)
• encrypted: Only members of A can decrypt, enforced cryptographically

This algebra enables rigorous, composable modeling of trust, replication, and control in distributed, encrypted systems.